Teaching Statement
Philosophy, experience, and courses — bridging space physics and data science in the classroom
My teaching philosophy is grounded in the conviction that science is learned by doing, not by watching. The most effective learning happens when students are confronted with real data and real uncertainty — the messiness of actual science — rather than textbook problems with clean answers.
In space physics and data science alike, we work with noisy, incomplete, multi-source datasets that require both domain intuition and quantitative rigor to interpret. My goal is to equip students with both: a physical intuition for why electromagnetic waves behave as they do, why the ionosphere changes with latitude and time, why machine learning models fail in ways their training metrics don't reveal — and the computational tools to investigate these questions independently.
"A student who can ask a better question is more valuable than one who can answer a standard exam. My job is to cultivate question-askers."
I structure courses around three pillars: (1) conceptual foundations built through physical reasoning and intuition-building exercises; (2) computational practice using Python, real datasets, and reproducible workflows; and (3) scientific communication — writing, presentation, and peer critique. These pillars reflect the skills students actually need in research and industry.
Graduate Research Mentor
Supervised and mentored multiple graduate students and undergraduate research assistants in SuperDARN data analysis, Python software development, and space weather simulation workflows. Co-mentored one Ph.D. student and two M.S. students on solar flare and geomagnetic induction research.
Google Summer of Code Mentor
Mentored an undergraduate student contributor to the SuperDARN Python visualization toolchain (pyDARN). Guided the student through open-source development practices, code review, documentation, and publication standards.
Outreach Volunteer — Pre-College Programs
Volunteer faculty in Summer Camp programs for pre-college students: Imagination (rising 7th–8th graders) and Pathways for Future Engineers (first-generation rising 10th–12th graders). Led sessions on space weather, electromagnetic phenomena, and data visualization using Python.
Science & Collaborative Talks — Freshman Outreach
Participated in collaborative seminars for incoming engineering freshmen, speaking on research opportunities, the path from coursework to research, and the value of data-driven thinking across engineering disciplines.
Conference Judging & Student Mentoring
Outstanding Student Presentation Award (OSPA) judge at AGU Fall Meeting since 2021. Session convener at AGU 2022. CEDAR student paper judge (2022, 2024). NASA program reviewer (2023). NSF program reviewer (2024). These roles allow me to support and evaluate the next generation of space weather researchers.
Space Physics & Space Weather
| Course Title | Level | Key Topics |
|---|---|---|
| Introduction to Space Weather | Undergrad/Grad | Solar activity, geomagnetic storms, ionospheric physics, space weather impacts on technology |
| Ionospheric Physics | Graduate | D/E/F-region chemistry, photochemistry, electrodynamics, radar remote sensing, TIDs |
| Space Weather Instrumentation | Graduate | SuperDARN HF radars, magnetometers, ionosondes, satellite data processing, experimental design |
| Electromagnetic Wave Propagation | Undergrad/Grad | Maxwell's equations, plane waves, reflection/refraction, HF propagation, plasma electromagnetics |
Data Science & Analytics
| Course Title | Level | Key Topics |
|---|---|---|
| Scientific Python for Geoscience | Undergrad/Grad | NumPy, Pandas, Matplotlib, xarray; time series analysis; NetCDF/HDF5 data; reproducible research |
| Machine Learning for Physical Sciences | Graduate | Supervised/unsupervised learning; neural networks; probabilistic ML; scikit-learn, TensorFlow; physics-informed ML |
| Applied Statistics for Scientists | Undergrad/Grad | Hypothesis testing, regression, Bayesian inference, uncertainty quantification, experimental design |
| Complex Systems Analysis | Graduate | Nonlinear dynamics, time series analysis, network theory, spectral methods, cross-disciplinary applications |
📊 Data-First Learning
Every concept is illustrated with real observational data. Students work directly with SuperDARN data, satellite observations, and open geophysics datasets from day one — building the habit of questioning data quality before drawing conclusions.
🐍 Open-Source Tooling
All courses use Python and open-source tools. I believe proprietary software creates unnecessary barriers to entry and reproducibility. Students graduate with skills immediately applicable in research, industry, and beyond.
🔁 Iterative Problem-Solving
Projects are designed with multiple checkpoints — hypothesis, exploration, analysis, communication. Students learn that research is an iterative process of refinement, not a linear march from question to answer.
🌍 Broad Impact & Context
I connect abstract physics to tangible real-world impacts: GPS degradation, power grid vulnerability, internet cable risk. Students should leave understanding why the science matters beyond the classroom.
I am committed to making science accessible and welcoming to students from all backgrounds. My outreach work with CEED at Virginia Tech has focused specifically on first-generation and underrepresented students in engineering. I bring this commitment into the classroom through active learning strategies that reduce reliance on prior privilege (such as well-resourced high school backgrounds), through explicit discussion of how science is done and who does it, and through mentoring that extends well beyond office hours.
My own path — from Kolkata, India, through TCS Innovation Labs, to a Ph.D. at Virginia Tech, to leading NASA-funded research — is a reminder that neither geography nor starting conditions determine scientific potential. That perspective informs how I teach.